1. Field of the Invention
The invention relates to vapor-phase heterogeneous reactions. More particularly, it relates to the dispersing of gases and solids in a liquid phase, reacting the vapor-phase components in the presence of a solid catalyst, and recovering the desired reaction product from either the liquid or vapor streams exiting the reactor.
2. Description of the Prior Art
Heterogeneous vapor-phase reactions are used to produce many large volume organic and inorganic chemicals. Examples of vapor-phase processes include oxidation, ammoxidation, and oxychlorination of olefins, alkanes, and inorganic species to yield a variety of chemicals, e.g. acrylic acid, acrylonitrile, ethylene dichloride, ethylene oxide, hydrogen peroxide, maleic anhydride, methanol, phthalic anhydride, propylene oxide, vinyl acetate and formaldehyde. In commercial processes, these vapor-phase reactions are conducted in fixed-bed or fluidized-bed (two-phase) reactors.
Although no two processes are exactly alike, commercial versions of vapor-phase processes typically possess at least one characteristic, i.e., a dilute reaction stream. The actual volume fraction of the process stream entering the reactor that is converted to products is typically small. The remainder of the stream consists of inerts and unconverted reactants. Consequently, process equipment associated with reactor process stream, the recycle stream, and product recovery and purification streams must be sized to accommodate the combined flow of reacting and nonreacting species.
The excess diluent represents a process "inefficiency" that adds to fixed capital and operating costs associated with the processes. The requirement for excess diluent arises from several process constraints inherent in conventional fixed-bed and fluidized-bed vapor-phase reactor operations. These constraints include:
1. Limitations due to inadequate rates of heat transfer to remove heat of reaction from catalyst particles:
Vapor-phase oxidations are typically highly exothermic reactions. Reaction rates are limited by the rate at which heat can be removed from the solid catalyst. Rapid reaction rates or poor rates of heat transfer result in generation of catalyst hot spots, which reduce product yields and catalyst lifetimes.
2. Limitations due to conversion-selectivity trade-offs:
Frequently, in heterogeneous vapor-phase oxidations, as the extent of reaction, i.e., conversion, increases, the catalyst selectivity to the desired product decreases. Loss in selectivity is often due to side and secondary reactions forming undesirable by-products, e.g. carbon dioxide. In economic terms, these trade-offs balance effectiveness of raw materials utilization and, in some cases, the size of recycle streams.
3. Limitations in feed compositions due to flammability and explosion hazards:
Flammable compositions of mixtures containing oxygen and hydrocarbons or other combustibles are demarcated by upper and lower flammability limits. Fixed-bed vapor-phase oxidation processes are typically designed to ensure that the compositions of all process streams lie outside the flammable region to avoid explosion hazards.
These limitations are particularly important in fixed-bed processes. In fluidized-bed processes, heat removal from the catalyst is much improved over fixed-bed designs, and explosion risks are considerably reduced by separating the organic and air feed streams and operating without recycle. Fluidized- bed processes typically run at very high conversion of organic reactant and relatively low conversion of oxygen. However, like their fixed-bed counterparts, they suffer from excess inerts in the reactor process stream. Since oxygen efficiency is sacrificed to achieve high conversion of organic reactant, these processes can not tolerate the relatively high cost of the needed oxygen, and instead use feed air as the oxygen source.
A new reactor development that eliminates heat transfer limitations, reduces or eliminates the need for excess inerts, and ensures safe operation could provide significant savings in the capital equipment and operating costs associated with chemical manufacture.
There are two classes of prior art relating to vapor phase reactions, one related to conventional two-phase reactor systems, and the other related to three-phase reactor systems. The two-phase systems cannot fully address all three issues raised above.
In general, two-phase reactor designs solve the problems of flammability and heat removal by dilution of the reaction stream. By utilizing inert components or operating at low conversions with large recycle streams, the composition of the reaction stream can be maintained outside the flammable region. This dilution using inerts and unconverted reactants in the recycle stream also serves to mitigate catalyst heating effects by reducing the heat of reaction per unit volume of feed.
Heat transfer constraints can also be significantly reduced by utilizing fluidized-bed designs in preference to fixed-bed designs. Smaller catalyst particle size and higher solid-vapor heat transfer coefficients associated with fluidized-bed designs both contribute to the improved performance. However, trade-offs also pertain with respect to fluidized bed designs. Backmixing, which can lead to reduced catalyst selectivity, often results in lower reaction yields and higher rates of production of undesired reaction products.
Three-phase reactors offer significantly better heat transfer than is obtainable using two-phase reactors. Higher rates of heat removal are achieved by contacting the catalyst directly with a liquid solvent.
Three-phase reactors come in several forms: (a) fixed beds, either trickle-bed or bubble-bed reactors, depending on whether the vapor phase is continuous or not; (b) ebullated-bed reactors, the three-phase equivalent to a fluidized-bed reactor; and (c) slurry reactors in which the catalyst and vapor phases are dispersed in the liquid phase, either with or without forced convection. The predominant application of three-phase reactors has been to liquid-phase hydrogenations, although some liquid-phase oxidations have also been proposed.
In three-phase systems, the distinction between vapor-phase and liquid-phase heterogeneous catalysis is not entirely obvious since both phases coexist with the solid catalyst. It is possible, however, to distinguish between vapor-phase and liquid-phase catalytic processes in these three-phase systems, based on the mass transfer mechanisms and the resultant reaction rates. If the vapor-phase reactants must diffuse through the liquid to reach the active catalyst sites, the overall reaction rates will be much slower than if mass transfer proceeds through the vapor phase. Thus, liquid-phase processes are characterized by much lower space velocities and slower reaction rates than vapor-phase processes.
Several three-phase, liquid-phase oxidation processes using fixed-bed and slurry reactor configurations are disclosed in the prior art. These include, for example, carbonylation (Drent, U.S. Pat. No. 5,149,868), direct oxidation of hydrogen to hydrogen peroxide (Hiramatsu et al., U.S. Pat. No. 5,132,099) aqueous-based oxidations and reductions (Mori et al., U.S. Pat. No. 4,578,112), and olefinic oxidation to epoxides (Shingu, U.S. Pat. No. 2,985,668 and Bobolev et al. U.S. Pat. No. 3,957,690). These processes remove the heat transfer limitations, and in many cases, provide yields and selectivities superior to commercially practiced technology. They have failed to achieve commercial significance because, as liquid-phase processes, they do not generate commercially acceptable rates of reaction.
Reaction rates comparable to vapor-phase processes have been achieved recently in three-phase reactors by using non-wetted support material. On such supports, the solvent does not capillary condense in the pores of the support, and the reactants have direct access to the active catalyst sites by vapor-phase transport without the additional mass transfer resistances associated with absorption and diffusion in the liquid phase.
Non-wetted catalyst supports are well-known in the literature. Initially developed for catalytic exchange of deuterium between hydrogen gas and water by Stevens, Canadian Patent No. 907292, and Butler et al., U.S. Pat. No. 4,228,034, and for redox reactions in aqueous systems by Shkol'nikov et al., Kinetica i Kataliz 20, 768-772 (1979), support materials were treated with teflon to make them hydrophobic. Virtually any conventional catalyst support can be rendered hydrophobic by such treatment, e.g. silicas, aluminas, titanium dioxide, and others. Additionally, a variety of other hydrophobic supports including fluorinated carbons, porous polymeric resins, and high silica zeolites have been disclosed in the literature, e.g. Rolston et al. Catalysis on the Energy Scene, and Kaliaquine and Mahey, Eds., Elsevier, 1984.
Hydrophobic, i.e., non-wetted, supports have been tested in three-phase trickle-bed reactors and exhibit vapor-phase reaction characteristics. Some representative examples include teflon-treated activated carbon catalyst supports used in fixed-bed reactors to oxidize SO.sub.2 in aqueous solutions (Goto and Morita, Chem. Eng. Commun. 60, 253-259, 1987). A fluorinated carbon is employed in a trickle-bed reactor used to oxidize organic contaminants in industrial wastewaters (Chuang et al. Ind. Eng. Chem. Res. 31, 2466-2472, 1992). Several different supports, including a styrene divinylbenzene copolymer, a fluorinated carbon, and silicalite, are used as supports for direct partial oxidation of olefins to carboxylic acids (Chuang and Fu, U.S. Pat. No. 5,210,319).
For purposes hereof, the term "non-wetted" as applied to a catalyst support/solvent combination is determined by adsorption and capillary condensation behavior, and not the applicable contact angle. The only requirement of the support is that the active component must be accessible to the reactants via transport through the vapor phase alone, i.e., at reactor conditions, the solvent must not block the access of the reactants by condensing in the pores of the support.
The three-phase reactor processes described in the prior art provide better heat transfer than conventional vapor-phase processes due to the higher heat capacity of the liquid phase. They also provide for in situ product recovery in the liquid phase. However, there are two significant problems with current three-phase reactor approaches as applied to vapor phase heterogeneous catalysis.
The first problem in inherent in trickle-bed (fixed-bed) designs utilizing hydrophobic catalyst supports in vapor-phase oxidation processes. The catalyst support must be hydrophobic to ensure vapor-phase reaction rates. It must also be intimately contacted with liquid to remove the heat and recover reaction products. For example, in trickle-bed reactors, the support catalysts are bonded to ceramic packing. A solvent flows through the column with the gas and contacts the catalyst and removes the oxidation product and the heat of reaction. However, if the support is hydrophobic, liquid flow will tend to be characterized by liquid rivulets with large portions of the catalyst bed remaining dry. Consequently, heat and mass transfer rates will be low, resulting in catalyst overheating and inefficient product recovery.
The second serious limitation of this approach is that such trickle-bed processes are co-continuous in the liquid and vapor phases. Consequently, there is significant risk of explosion if operating within the flammability limits. To eliminate explosion hazards, such processes, as described in the prior art, use excess reactant or inert species to dilute the reactive species and minimize the risk of explosion.
There remains in the art, therefore, a need for further development in three-phase reactor processes and systems. Such development will desirably overcome the two significant problems referred to above with respect to vapor-phase heterogeneous catalysis.
It is an object of the invention, therefore, to provide an improved reactor system and process for the carrying out of vapor phase heterogeneous reactions.
It is another object of the invention to provide an improved reactor system and process for the carrying out of non-wetted solid catalyst, vapor-phase heterogeneous reactions.
With these and other objects in mind, the invention is hereinafter described in detail, the novel features thereof being particularly pointed out in the appended claims.